1,292,693 / omeprazole formulation / LOSEC
In the accounting of profits portion of this bifurcated trial,1 the parties settled most issues, but required determination of four specific points. The first, discussed in this post, is whether Apotex had a non-infringing alternative available during the period of infringement. The decision raises two significant issues. One is whether an infringer must know at the time of the infringement that the NIA is non-infringing, for it to be considered available. More generally, this case is novel in that Apotex attempted to establish case an NIA “defence”2 based solely on alternatives which had been developed only for the purposes of litigation [24], [92]. This contrasts with the leading cases to date, Lovastatin FCA 2015 FCA 171 and Perindopril FCA 2017 FCA 23. In Lovastatin (see here) the putative NIA had actually been developed and commercialized by Apotex prior to the infringement; in Perindopril (here) the issue involved export of infringing product made in Canada, and the NIA was that the product would have been exported to the same countries, and made by the same process in other jurisdictions where it was not patented. In the leading US decision, Grain Processing 185 F.3d 1341 (Fed. Cir. 1999), the alternative did not actually exist at the date of infringement, but was commercialized almost immediately after infringement was established. In this case, in contrast, the putative NIAs were seven alternative formulations which Apotex never commercialized and has no intention of ever commercializing [24]. They were developed solely for the purpose of arguing that they could and would have been commercialized in the but for world.
Apotex won on the law, but lost on the facts. Barnes J held that these types of alternatives could in principle be considered, but he found that Apotex failed to establish that the particular alternatives it developed would have been commercializable, or that they would have received regulatory approval. While this ultimately turned on the specific facts of the case, there is perhaps a broader lesson. The testing of the alternative formulations that was done in this case fell far short of what would be required for regulatory approval. The tests used to predict both bioequivalency and stability were appropriate for use in a formulation screening exercise, but this was far from enough to persuade Barnes J that the formulations would probably have been approved and commercially successful [139], [146]. This suggests that while post-infringement NIAs are acceptable in principle, it will be difficult for a generic to surmount the necessary evidentiary hurdles to establish the NIA on the facts. Apotex also failed to establish it would have sourced non-infringing product from a third party (which I will discuss tomorrow) [196]. On the whole, this decision should give comfort to patentees who were concerned that the acceptance of the NIA defence would open the floodgates to speculative defences.
On the law, AstraZeneca made three arguments of principle. First, relying on MacKay J’s decision in Wellcome v Apotex (TMP) (1998), 82 CPR (3d) 466, [32]-[33}, it argued that a valid NIA must be perceived or known by the infringer to be non-infringing at the point of the infringement [12]. [13]. To simplify slightly, in TMP the final product, TMP, could be made by two processes, both of which were patented by Wellcome. At trial, one of the process patents was held to be valid and infringed, but the other was held to be invalid. Apotex argued that it – or more precisely, its suppliers – could have used the process protected by the invalid patent. MacKay J rejected this argument, saying “there is really no evidence that at the time the product was acquired there was available product then known to be non-infringing” [33]. In part this was because the alternative patent was not then known to be invalid, and in part because Apotex did not know which process was used by its suppliers. On the whole, though the discussion is brief and somewhat cryptic, TMP does seem to me to support AstraZeneca’s argument. Barnes J distinguished TMP on the basis that MacKay J also considered whether Apotex “could have known” [14]. I don’t find this distinction entirely persuasive. MacKay J’s point, as I read it, is that not only did Apotex not actually know, it might not have been possible for it to know – this reinforces the importance of knowledge, rather than diminishing it. In any event, Barnes J did accept that knowledge is relevant, saying “[k]nowing whether or not a proposed NIA would infringe is, of course, a factor in determining whether the infringer ‘would have’ employed it in place of the infringing product” [14], though “this falls well short of making prior knowledge of non-infringement an absolute pre-requisite to the assertion of a NIA” [13]. Thus, Barnes J agreed with MacKay J that knowledge that the NIA is non-infringing is relevant, and the difference between them, if any, is whether it should be a complete bar.
But why should it be relevant at all? As Barnes J noted, it apparently follows from the “would have” branch of the test for availability, evidently on the notion that the infringer – eg Apotex in TMP – would not have known to insist that its suppliers use the non-infringing alternative because it did not know that it was non-infringing. But this seems dubious in principle, as this approach allows an invalid patent to shield behind a valid one. To go back to first principles, the reason alternatives are considered is that the return to a patentee should be commensurate with the value of the patented technology, which is equal to its value as compared with the best non-infringing alternative: Lovastatin FCA [56]. A technology may be innovative and patentable, yet not valuable, if the product is no better or cheaper than a mundane alternative. If the infringer must know that a patent is invalid before that technology can be used as an alternative, the patentee can improperly inflate the apparent value of the innovative technology by patenting obvious technologies that serve the same end, even though those patents are ultimately invalidated. This is a kind of remedial evergreening that is not addressed by the usual expedient of invalidating the offending patents. It will be interesting to see how the law continues to develop on this point.
AstraZeneca also argued that an NIA “must be ‘foreseeable’ to the infringer at the relevant time” [13], citing Lovastatin FCA [93]-[94]. Barnes J dismissed this as a “stray reference” [15], saying
If foreseeability meant that the infringer must have the asserted NIA in mind at the time
of the infringement, it could potentially punish those who had no idea their product was
infringing while rewarding those who had an appreciation of the risk and courted it, but
nevertheless had a back-up, work-around solution available.
Barnes J’s observation strikes me as compelling.
Finally, citing Grain Processing 185 F 3d 1341 (Fed Cir 1999) and Micro-Chemical 318 F 3d 1119 (Fed Cir 2003), AstraZeneca argued that an NIA “requiring the infringer to ‘invent around the patented technology’ is not considered to be ‘available’ to the infringer” [16]. Barnes J rejected this, saying that cases stand only for the proposition that “[t]he time and effort of coming up with a non-infringing solution is certainly relevant to whether the infringer would have pursued it, but they are not absolute barriers to the defence” [16]. That conclusion strikes me as both a correct reading of those cases, and sound as a matter of principle.
In conclusion, Barnes J held that as a matter of law:
in the hypothetical, but for pharmaceutical world the infringer’s failure to produce a
viable NIA formulation in the real world is not a threshold bar to the use of the NIA
defence . [18]
The correct question is simply “Could the infringer have made the product had it attempted to do so at the relevant time and would the infringer have sold the product on some reasonable financial basis in substitution for the infringing product?” [18].
While the law was largely decided in Apotex’ favour, Barnes J pointed out that post-infringement development of an NIA may present serious problems of proof, which were manifest in this case [20].
One general point is that the Court should be wary of hindsight bias when assessing the availability of post-infringement development of an NIA [21], in the same way as the Court must be wary of hindsight bias was assessing an obviousness attack.
But this case did not turn on hindsight bias. It turned on more basic evidentiary problems:
[24] When a pharmaceutical NIA has been created and has obtained regulatory approval,
one is not left to wonder whether it “could” have been available for use (assuming a
capacity to obtain it in commercial amounts). In this case, however, Apotex’s self-created
NIAs were made in non-commercial batches, without full stability, bioequivalency or
clinical studies, and without obtaining the required regulatory approvals for commercial
use.
Nontheless, Barnes J did not dismiss Apotex’ tests out of hand, but at the end of the day, the data presented by Apotex was flawed and speculative in many respects: “shortcuts like the ones employed by Apotex may be acceptable in the early stages of product selection, but they are entirely unacceptable in the context of proving the viability of a NIA for the purpose of this case” [143]. For much the same reasons, Apotex could not establish that its proposed alternatives would have received regulatory approval [144]-[152].
1 The liability portion is Apotex Inc v AstraZeneca Canada Inc 2017 FCA 9 var’g 2015 FC 322 as amended by 2015 FC 671: see FCA Claim Construction FCA Overbreadth FC Product Claim FC Foreign Issue Estoppel
2 I don’t like the term NIA “defence” because in most circumstances it does not serve as a complete defence, but only to reduce the quantum that would otherwise be awarded. Nonetheless, the term is convenient, and it seems to have become established, so I will use it, albeit with this caveat.
No comments:
Post a Comment